The overall objective of this study is to test the hypothesis that a true difference exists in fetal and neonatal loss and maternal smoking in some population groups but not in others because of differential interactions between maternal smoking and other risks. The data of the Ontario Perinatal Mortality Study provides the basis for the work to be performed. The study data comprise detailed, uniform records for all single births in ten teaching hospitals in Ontario for 1960 and 1961, a total of 51,490 births, including 701 fetal deaths and 655 early neonatal deaths (under 8 days). Analysis will include measurements of dose-response, interactions between smoking and other variables effecting fetal weight and survival, calculation of relative risks and of risks attributable to smoking; multivariate analysis as appropriate; analysis of relationships between smoking, placenta weight, duration of pregnancy, fetal weight, and mortality.